Statements (28)
Predicate | Object |
---|---|
gptkbp:instanceOf |
gptkb:model
|
gptkbp:affiliation |
gptkb:Seoul_National_University
|
gptkbp:application |
image super-resolution
|
gptkbp:architecture |
gptkb:convolutional_neural_network
|
gptkbp:citation |
over 3000
|
gptkbp:countryOfOrigin |
gptkb:South_Korea
|
gptkbp:fullName |
Enhanced Deep Super-Resolution Network
|
https://www.w3.org/2000/01/rdf-schema#label |
EDSR
|
gptkbp:improves |
gptkb:SRCNN
gptkb:VDSR |
gptkbp:input |
low-resolution image
|
gptkbp:introduced |
gptkb:Kyoung_Mu_Lee
gptkb:Sanghyun_Son gptkb:Seungjun_Nah Bee Lim Heewon Kim |
gptkbp:notableFeature |
removes batch normalization
uses residual blocks |
gptkbp:openSource |
yes
|
gptkbp:output |
high-resolution image
|
gptkbp:publishedIn |
gptkb:CVPR_2017
|
gptkbp:relatedTo |
gptkb:SRGAN
FSRCNN RCAN |
gptkbp:repository |
https://github.com/twhui/DIV2K_EDSR_Tensorflow
|
gptkbp:trainer |
DIV2K
|
gptkbp:bfsParent |
gptkb:VDSR
|
gptkbp:bfsLayer |
6
|